ABSTRACT Upifitamab rilsodotin—an antibody drug conjugate (ADC)—comprises a NaPi2b‐targeted antibody conjugated to an auristatin‐based payload (auristatin F‐hydroxypropylamide [AF‐HPA]). AF‐HPA is metabolized by cytochrome P450 3A4 (CYP3A4) and, to a lower extent, by CYP3A5 and demonstrates both reversible and time‐dependent inhibition of CYP3A4. AF‐HPA is also a P‐glycoprotein (P‐gp) substrate. A PBPK model was developed using a mixed “bottom‐up” and “top‐down” modeling approach with a combination of in vitro, nonclinical, and clinical ADME/PK data. The model recapitulated the clinical PK of conjugated and unconjugated AF‐HPA. Simulations were used to predict the potential of unconjugated AF‐HPA to be a victim or perpetrator of clinical drug–drug interactions (DDI) and predict the impact of hepatic impairment on the exposure to unconjugated AF‐HPA. Simulations suggested negligible potential for clinical DDI between unconjugated AF‐HPA and CYP3A substrates. Simulations also showed ~30% increase in unconjugated AF‐HPA exposure following an IV dose of 36 mg/m 2 in the presence of itraconazole, an inhibitor of both CYP3A4 and P‐gp. A negligible change in the exposure to unconjugated AF‐HPA was predicted in patients with mild hepatic impairment, which aligned with observed clinical data. The model predicted a ~1.5‐fold increase in unconjugated AF‐HPA AUC and negligible change in the C max in patients with moderate and severe hepatic impairment. Finally, this PBPK model may be applied (with modification to the conjugated drug sub‐model parameters) to predict DDI and hepatic impairment potential for other ADCs with the same linker and payload.

Background: The murine double minute 2 (MDM2) oncoprotein is a key E3 ubiquitin ligase that degrades the tumor-suppressor p53. Targeting of the MDM2/p53 interaction to stabilize p53 and induce apoptosis in wildtype (WT) p53 tumors is an emerging therapeutic approach in WT p53 hematologic and solid tumor malignancies. However, MDM2/p53 small molecule inhibitors (SMIs) induce a p53/MDM2 feedback loop, resulting in upregulation of MDM2 protein levels and p53 pathway inhibition thus limiting their biological activity and clinical application. KT-253 is a novel, highly potent heterobifunctional MDM2 degrader with >200-fold higher potency than MDM2 SMIs that can suppress p53-dependent MDM2 protein feedback upregulation. We have shown previously that because of its superior pharmacological profile, a single dose of KT-253 was sufficient to induce rapid apoptosis and sustained tumor regression, supporting an intermittent dosing schedule of KT-253. Aims: Here, we assessed how pulse dosing versus exposure matched fractionated doses of KT-253 drives efficacy in mouse xenograft models and characterized the underlying molecular mechanisms associated with the different dosing regimens. In addition, we compared the efficacy of KT-253 to the clinical equivalent dosing schedule of a p53/MDM2 SMI. Methods: Mice were treated with KT-253 on a 3-week dosing cycle, either with a “pulse” dose or various intermittent dosing regimens at doses matched for total AUC. MDM2 SMI DS-3032 was dosed at its clinically equivalent dose and regimen of 3 days on/11 days off. Pharmacodynamic effects were assessed at mRNA level by quantitative PCR and protein level by proteomics. Results: We show in the RS4;11 xenograft model that a pulse IV dose of KT-253 led to rapid and efficient MDM2 degradation, potent p53 induction, apoptosis and sustained tumor regression. In contrast, exposure matched fractionated doses of KT-253 administered on various frequent intermittent dosing schedules spanning a three-week cycle resulted in less efficient MDM2 degradation, modest activation of p53 downstream signaling, cell cycle arrest and tumor stasis. Furthermore, the p53/ MDM2 SMI DS-3032 led only to modest p53 activation and showed limited anti-tumor activity. The acute apoptotic mechanism of KT-253 observed with pulse dosing was confirmed across multiple cell lines and xenograft models. Our results suggest that acute and potent MDM2 degradation is necessary to induce irreversible apoptosis and tumor regression, which can only be achieved with a pulse dose of KT-253 and not with the repeat dosing regimens of MDM2/p53 SMIs. Summary/Conclusion: In summary, our data suggest that pulsatile, higher doses of the MDM2 degrader KT-253 are superior to more frequent lower doses by inducing rapid apoptosis in MDM2-dependent cancer cells. The acute pulse dosing regimen of KT-253 has the potential to demonstrate an improved efficacy and safety profile compared to the more frequent dosing of MDM2/p53 SMIs in the clinic. Keywords: Acute lymphoblastic leukemia, p53, Acute myeloid leukemia

Supplementary Figure from Pharmacologic Activation of STING in the Bladder Induces Potent Antitumor Immunity in Non–Muscle Invasive Murine Bladder Cancer

The murine double minute 2 (MDM2) oncoprotein is a key E3 ubiquitin ligase that degrades and thereby inactivates the tumor-suppressor p53. Targeting of the MDM2/p53 interaction with reversible small molecule inhibitors (SMI) to stabilize p53 and induce apoptosis in wildtype (WT) p53 tumors has been an emerging therapeutic approach in AML and in other WT p53 hematologic and solid tumor malignancies. However, recent clinical trials with MDM2 inhibitors, especially in R/R AML, have resulted in suboptimal clinical activity, highlighting the need for novel therapeutic approaches. KT-253 is a novel, highly potent heterobifunctional MDM2 degrader that suppresses p53-dependent MDM2 protein upregulation that is known to be triggered by the MDM2 SMIs and thereby limits their clinical activity. Previously, we have shown that KT-253 has superior activity compared to MDM2 SMIs, demonstrating >200-fold improvements in both in vitro cell growth inhibition and apoptosis. Because of its superior pharmacological profile, a single dose of KT-253 was sufficient to induce rapid apoptosis and sustained tumor regression in the MV4;11 AML and RS4;11 ALL cell line-derived (CDX) mouse xenograft models, supporting an intermittent dosing schedule of KT-253. Because of the initial promising clinical activity of SMIs in AML, and the potential for KT-253 to provide superior activity, we assessed the activity of KT-253 in AML patient-derived xenograft (PDX) models. In vivo, KT-253 administered once every 3 weeks at 1 mg/kg led to tumor regression in a variety of these models, including CTG-2227 (Figures 1A and B). In addition, in an AML CDX model resistant to venetoclax such as MOLM13, KT-253 administered in combination with venetoclax once every three weeks showed more substantial activity than the single agents alone and led to sustained tumor regression. In summary, intermittent dosing of the clinical candidate KT-253 achieved tumor regression in a variety of AML PDX and CDX models. Combination of KT-253 with the AML standard of care treatment venetoclax achieved durable tumor regression in additional AML xenograft models that are resistant to standard of care, suggesting potential benefit to an expanded patient population. In addition, we have identified hematological and solid tumor indications that respond acutely to MDM2 degradation, we plan to share more details at upcoming meetings. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Signal transducer and activator of transcription 3 (STAT3), an “undruggable” transcription factor activated by a variety of receptor- and non-receptor tyrosine kinases, plays a critical role in activation pathways triggered by cytokines, hormones, and growth factors, which makes it an attractive target for the treatment of autoimmune and autoinflammatory disorders. Kymera has developed heterobifunctional molecules that selectively target STAT3 for degradation by the ubiquitin-proteasome pathway. These degraders have broad and potent activity in-vitro against TLR receptor and cytokine-induced activation of immune and stromal cells and attendant mediators release such as MCP-1 (CCL2) and collagen1a1. STAT3 degradation in CD4+ T cells potently inhibited Th17 development, decreasing IL-17, IL-22, IL-8/CXCL8, and TNFa production, with concomitant increase in Treg numbers, that was superior to JAK1/2 kinase inhibition. STAT3 degradation was subsequently evaluated in-vivo in murine models of autoimmune disease. In the Th17-driven inflammatory model of MOG-induced Experimental Autoimmune Encephalomyelitis (EAE), dose-dependent decrease of incidence, disease onset, clinical scores, and histopathology were observed in comparison to a S1P1 inverse agonist or steroid treatment. Ex-vivo MOG-stimulated cytokine release by leukocytes isolated from draining lymph node was also robustly inhibited. These data demonstrate the broad activity of STAT3 degradation in alleviating autoimmune inflammation in systems relevant to human disease.

Signal Transducer and Activator of Transcription 3 (STAT3) plays important roles in the transduction of signals from growth factors and cytokines in both normal and malignant cells. Upon activation, STAT3 controls expression of genes that regulate cell growth, survival, differentiation, stemness and cell-cell interactions. Aberrant activation of STAT3 has been observed in many cancers including lymphoma and leukemias through activating mutations, hyper-signaling through upstream regulators or loss of negative feedback regulation. Additionally, STAT3-mediated cross-talk in the tumor microenvironment results in suppression of immune surveillance compromising anti-tumor immunity.

The predictive performance of physiologically‐based pharmacokinetics (PBPK) models for pharmacokinetics (PK) in renal impairment (RI) and hepatic impairment (HI) populations was evaluated using clinical data from 29 compounds with 106 organ impairment study arms were collected from 19 member companies of the International Consortium for Innovation and Quality in Pharmaceutical Development. Fifty RI and 56 HI study arms with varying degrees of organ insufficiency along with control populations were evaluated. For RI, the area under the curve (AUC) ratios of RI to healthy control were predicted within twofold of the observed ratios for > 90% (N = 47/50 arms). For HI, > 70% (N = 43/56 arms) of the hepatically impaired to healthy control AUC ratios were predicted within twofold. Inaccuracies, typically overestimation of AUC ratios, occurred more in moderate and severe HI. PBPK predictions can help determine the need and timing of organ impairment study. It may be suitable for predicting the impact of RI on PK of drugs predominantly cleared by metabolism with varying contribution of renal clearance. PBPK modeling may be used to support mild impairment study waivers or clinical study design.

Background Signal Transducer and Activator of Transcription 3 (STAT3), a multifaceted transcription factor, is aberrantly activated across a variety of malignancies; however, its selective targeting has to-date remained a therapeutic challenge. STAT3 plays a pivotal role in shaping the tumor immune landscape through cancer cell-intrinsic mechanisms, direct regulation of immune cell function and via cancer cell- tumor microenvironment (TME) crosstalk, that collectively result in an immunosuppressive TME. Targeted protein degradation represents a novel therapeutic modality enabling direct targeting of previously undruggable oncoproteins. We have developed potent and selective STAT3 heterobifunctional degraders demonstrating activity across diverse tumor and immune cell types. Methods We investigated the immunomodulatory impact of STAT3 degradation on tumorigenesis in syngeneic mouse models representing cancers with heterogeneous immune milieus. Methods included in vivo pharmacological approaches, immunophenotyping and gene expression profiling. Results Treatment of CT-26 (colorectal cancer) and A20 (B-cell lymphoma) tumor-bearing mice with a STAT3 degrader resulted in significant tumor growth inhibition compared to controls, with loss of STAT3 protein in both tumor cells and TME. This was accompanied by a decrease in M2 polarized macrophages and concomitant increases in M1 polarized macrophages and tumor infiltrating lymphocytes. The anti-tumor responses were abrogated by antibody mediated CD8+ T cell depletion or by using immunodeficient host-strains implicating the observed efficacy to be predominantly driven by immune-directed mechanisms. Gene expression profiling of STAT3 degrader-treated CT-26 tumors showed marked increases in proinflammatory genes including T cell and M1 macrophage activation markers, compared to controls. Notably, induction of an Ifnγ-responsive gene signature (Ifnγ, Stat1, Cxcl9, Cxcl10, Ido1) suggested that STAT3 degradation results in a T-cell inflamed phenotype associated with responsiveness to immune checkpoint therapy (ICT). Furthermore, on-treatment tumors showed an upregulation of genes such as Pdl1, Ctla4, Lag3 which reflect T cell activation as well as counterregulatory mechanisms. Therefore, we evaluated STAT3 degradation in combination with anti-PD1 in these models which are poorly responsive to anti-PD1 monotherapy. Robust synergy was observed in the CT-26 model with 60% complete responses and development of immunological memory as confirmed by tumor re-challenge studies. Studies are underway to ascertain the applicability of this combination therapy in different tumor-immune contextures and indications, and to elucidate the mechanistic basis of synergy. Conclusions STAT3 degradation remodels an immunosuppressed TME activating anti-tumor immunity as monotherapy and effectively combines with anti-PD1. These data provide a rationale for selectively degrading STAT3 as a strategy to sensitize cancers with relevant immune contextures to ICT in the clinic.

E7766 represents a novel class of macrocycle-bridged dinucleotides and is under clinical development for immuno-oncology. In this report, we identified mechanism of systemic clearance E7766 and investigated the hepatobiliary transporters involved in the disposition of E7766 and potential drug interactions of E7766 as a victim of organic anion-transporting polypeptide (OATP) inhibitors. In bile-duct cannulated rats and dogs, E7766 was mainly excreted unchanged in bile (>80%) and to a lesser extent in urine (<20%). Sandwich-cultured human hepatocytes (SCHHs), transfected cells, and vesicles were used to phenotype the hepatobiliary transporters involved in the clearance of E7766. SCHH data showed temperature-dependent uptake of E7766 followed by active biliary secretion. In vitro transport assays using transfected cells and membrane vesicles confirmed that E7766 was a substrate of OATP1B1, OATP1B3, and multidrug resistance-associated protein 2. Phenotyping studies suggested predominant contribution of OATP1B3 over OATP1B1 in the hepatic uptake of E7766. Studies in OATP1B1/1B3 humanized mice showed that plasma exposure of E7766 increased 4.5-fold when coadministered with Rifampicin. Physiologically based pharmacokinetic models built upon two independent bottom-up approaches predicted elevation of E7766 plasma exposure when administered with Rifampicin, a clinical OATP inhibitor. In conclusion, we demonstrate that OATP-mediated hepatic uptake is the major contributor to the clearance of E7766, and inhibition of OATP1B may increase its systemic exposure. Predominant contribution of OATP1B3 in the hepatic uptake of E7766 was observed, suggesting polymorphisms in OATP1B1 would be unlikely to cause variability in the exposure of E7766. SIGNIFICANCE STATEMENT: Understanding the clearance mechanisms of new chemical entities is critical to predicting human pharmacokinetics and drug interactions. A physiologically based pharmacokinetic model that incorporated parameters from mechanistic in vitro and in vivo experiments was used to predict pharmacokinetics and drug interactions of E7766, a novel dinucleotide drug. The findings highlighted here may shed a light on the pharmacokinetic profile and transporter-mediated drug interaction propensity of other dinucleotide drugs.

Abstract Introduction: Calmette-Guerin (BCG)-unresponsive non-muscle invasive bladder cancer (NMIBC) is a significant unmet medical need. We report pharmacological, pharmacodynamic and pharmacogenomic characterization of E7766, a novel STING (stimulator of interferon genes) agonist, as a potent immunotherapy in BCG-insensitive orthotopic murine NMIBC models via intravesical administration. To our knowledge, this is the first demonstration of activity of a STING agonist in NMIBC tumor model. Methods: E7766 was examined for antitumor activity in two orthotopic murine NMIBC tumor models via intravesical administration and comparatively characterized in STING activation activity in primary human PBMC cells with defined STING genotypes. Local and systemic pharmacodynamic biomarker changes in tumor-bearing model were identified. Results: E7766, from a structurally novel class of macrocyclic STING agonists, showed specific and potent agonist activity in both human and mouse STING. In human PBMCs, E7766 demonstrated potent and consistent activity across seven tested human STING genotypes. By contrast, a reference cyclic dinucleotide STING agonist showed substantial variability, and was inactive in the REF-REF genotype. Intravesical administration of E7766 to two orthotopic murine models for BCG-insensitive NMIBC showed a dose-dependent and curative activity without serious adverse effects. In contrast, anti-PD1 treatment was ineffective in both models. Tumor-free animal derived from E7766 treatment completely rejected rechallenge of the same tumor cells, demonstrating establishment of effective tumor-specific memory immune response in those compound-treated animals. The anti-tumoral activity was accompanied with activation of IFN pathway, T cell infiltration, NK activity and antigen presentation in bladder wall, and robust induction of IFNβ, CXCL10 and other downstream effectors of STING activation inside the bladder cavity and in urine. Notably, the IFNβ gene induction was found in bladder, but not in blood from treated animals indicative of a primarily local STING activation by intravesical E7766, which was consistent with low systemic bioavailability of the compound via the administration route. Conclusions: These preclinical studies demonstrated a potent anti-tumor activity and induction of tumor-specific memory response by intravesically administered STING agonist E7766 in orthotopic murine models for BCG- and anti-PD1-insensitive NMIBC. A clinical study for intravesical E7766 in NMIBC patients has been initiated (NCT04109092) in North America. Citation Format: Kuan-Chun Huang, Chi Zhang, Kun Yu, Dae-Shik Kim, Vaishali Dixit, Renee Hukkanen, Hyeong-Wook Choi, Janna Hutz, Frank Fang, Xingfeng Bao. Demonstration of E7766, a novel STING agonist, as a potent immunotherapy in BCG-insensitive non-muscle invasive bladder cancer models via intravesical administration [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 592.

Physiologically-based pharmacokinetic (PBPK) modeling has been extensively used to quantitatively translate in vitro data and evaluate temporal effects from drug-drug interactions (DDIs), arising due to reversible enzyme and transporter inhibition, irreversible time-dependent inhibition, enzyme induction, and/or suppression. PBPK modeling has now gained reasonable acceptance with the regulatory authorities for the cytochrome-P450-mediated DDIs and is routinely used. However, the application of PBPK for transporter-mediated DDIs (tDDI) in drug development is relatively uncommon. Because the predictive performance of PBPK models for tDDI is not well established, here, we represent and discuss examples of PBPK analyses included in regulatory submission (the US Food and Drug Administration (FDA), the European Medicines Agency (EMA), and the Pharmaceuticals and Medical Devices Agency (PMDA)) across various tDDIs. The goal of this collaborative effort (involving scientists representing 17 pharmaceutical companies in the Consortium and from academia) is to reflect on the use of current databases and models to address tDDIs. This challenges the common perceptions on applications of PBPK for tDDIs and further delves into the requirements to improve such PBPK predictions. This review provides a reflection on the current trends in PBPK modeling for tDDIs and provides a framework to promote continuous use, verification, and improvement in industrialization of the transporter PBPK modeling.

Background This phase 1 study examined the safety, maximum-tolerated dose (MTD) and antitumour activity of E7449, a novel PARP 1/2 and tankyrase 1/2 inhibitor. Methods E7449 was orally administered once daily in 28-day cycles to patients with advanced solid tumours (50–800-mg doses). Archival tumour samples from consenting patients were evaluated for the expression of 414 genes in a biomarker panel (2X-121 drug-response predictor [DRP]) found to be predictive of the response to E7449 in cell lines. Results Forty-one patients were enrolled (13 pancreatic, 5 ovarian, 4 each with breast, lung or colorectal cancer and 11 with other tumour types). The most common grade ≥3 treatment-related adverse event was fatigue ( n = 7, 17.1%). Five patients experienced a dose-limiting toxicity (fatigue, n = 4, 800 mg; anaphylaxis, n = 1, 600 mg) for an MTD of 600 mg. E7449 exhibited antitumour activity in solid tumours, including 2 partial responses (PRs), and stable disease (SD) in 13 patients, which was durable (>23 weeks) for 8 patients. In 13 patients, the 2X-121 DRP identified those achieving PR and durable SD. E7449 showed good tolerability, promising antitumour activity and significant concentration-dependent PARP inhibition following 50–800-mg oral dosing. Conclusion The results support further clinical investigation of E7449 and its associated biomarker 2X-121 DRP. Clinical trial registration www.ClinicalTrials.gov code: NCT01618136.

Introduction STING (stimulator of interferon genes) is an emerging target for cancer immunotherapy. 2’,3'-cGAMP, a natural cyclic dinucleotide (CDN) STING agonist, and its phosphorothioate analogs, have drawn broad attention as lead molecules for STING targeted drug discovery. These CDNs, however, lack efficacy in some common STING genotypes disproportionally represented in non-Caucasians. Moreover, such CDNs have not fully addressed liability in chemical/metabolic stability. Here we report our chemistry approach to control STING agonist conformation to enhance binding affinity across all common STING genotypes and broaden the therapeutic potential of such compounds. Methods Our SBDD approach started with analysis of the binding pocket and key protein-ligand interactions to prioritize a focused set of analogs for chemical synthesis. Systematic SAR was built upon in vitro assays for STING binding affinity and activation of STING genotypes. X-ray single crystal structures were established for STING and diverse analogs, in free and bound states, to provide structural insight for rational analog design. Results Structural modeling was refined to evaluate different binding modes and dynamic conformational changes in the STING-ligand interface. We observed that STING-bound CDNs had the two ancillary nucleobases specifically oriented in close proximity with parallel pi-pi stacking and discovered that covalently linking the nucleobases advantageously pre-organize the bioactive constrained conformation for enhanced STING affinity. Our discovery established a novel class of macrocycle-bridged STING agonists (MBSAs). E7766, a representative of Eisai MBSA platform, shows superior in vitro activity against all the major human STING genotypes over reference CDNs, most distinctly in STINGREF. E7766 co-crystal structures with STINGWT and STINGREF provide structural basis for the added benefit of the topological novelty. The macrocyclic linker bridging the top of nucleobases perturbs the STING lid loop conformation and create new and specific interactions with both genotypes. In twelve subcutaneous tumor models in immune competent mice, single intra-tumoral injections achieved either complete regression or significant tumor growth delay with no serious adverse effect. E7766 also shows excellent chemical and metabolic stability, presumably conferred by conformational rigidity of the unique macrocycle bridge. More biological characterization of E7766 can be found in abstract #. Conclusion Eisai successfully discovered E7766, a representative of a novel class of macrocycle-bridged STING agonist topologically distinct from conventional STING agonists. E7766 demonstrated pan-genotypic STING activation, potent anti-cancer activities and excellent chemical and metabolic stability for further development. Citation Format: Atsushi ENDO, Dae-Shik Kim, Kuan-Chun Huang, Ming-Hong Hao, Steven Mathieu, Hyeong-wook Choi, Utpal Majumder, Xiaojie Zhu, Yongchun Shen, Kristen Sanders, Thomas Noland, Dinesh Chandra, Yu Chen, Karen TenDyke, Kara Loiacono, Donna Kolber-Simonds, Rongrong Jiang, Vaishali Dixit, Janna Hutz, John Wang, Xingfeng Bao, Francis Fang, Nadeem Sarwar. Discovery of E7766: A representative of a novel class of macrocycle-bridged STING agonists (MBSAs) with superior potency and pan-genotypic activity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4456.

Physiology-based pharmacokinetic (PBPK) modelling has been extensively used to quantitatively translate in vitro data and evaluate temporal effects from drug-drug interactions (DDIs), arising due to reversible enzyme and transporter inhibition, irreversible time-dependent inhibition, enzyme induction, and/or suppression. PBPK modelling has now gained reasonable acceptance with the regulatory authorities for the Cytochrome-P450-mediated DDIs and is routinely used. However, the application of PBPK for transporter-mediated DDIs (tDDI) in drug development is relatively uncommon. Since the predictive performance of PBPK models for tDDI is not well established, here we represent and discuss examples of PBPK analyses included in regulatory submission (FDA, EMA, and PMDA) across various tDDI. The goal of this collaborative effort (involving scientists representing 17 pharmaceutical companies in the Consortium and from academia) is to reflect on the use of current databases and models to address tDDIs. This challenge the common perceptions on applications of PBPK for tDDI and further delve into the requirements to improve such PBPK predictions. This review provides a reflection on the current trends in PBPK modelling for tDDIs and provides a framework to promote continuous use, verification and improvement in industrialisation of the transporter PBPK modelling.

Introduction: We report discovery and characterization of E7766, a structurally novel STING agonist, as a potential immunotherapy for solid cancers through intratumoral (IT) administration and for Bacillus Calmette-Guerin (BCG) unresponsive non-muscle invasive bladder cancer (NMIBC) through intravesical (VE) administration. Methods: E7766 was designed and synthesized to optimize the potency of binding to dimerized STING proteins of different genetic isoforms. The compound was extensively and comparatively characterized in a variety of biochemical, molecular and cellular, in vivo, ex vivo, and primary human tumor and cellular studies for potency, mechanisms and translational biomarkers. Novel preclinical models to mimic orthotopic NMIBC and deep lesion metastasis were developed, and co-crystalization with recombinant proteins of genetic variations was performed. Results: E7766, a novel Macrocycle-Bridged STING Agonist, showed highly specific and potent agonist activity in both human and mouse STING. In human PBMCs, E7766 demonstrated potent and consistent activity across seven tested human STING genotypes (IC50, 0.15-0.79 μM). By contrast, a reference cyclic dinucleotide STING agonist showed weaker potency and substantial variability across genotypes (IC50, 1.88 μM - >50 μM). Co-crystal structures indicated a structural basis for the superior interactions of E7766 with STING proteins compared with conventional cyclic dinucleotide STING agonists. Intravesical administration of E7766 to a preclinical orthotopic mouse bladder cancer model mimicking the BCG-unresponsive NMIBC demonstrated a dose-dependent and curative activity without serious adverse effects. The anti-tumoral activity was associated with a robust induction of IFNβ, CXCL10 and other downstream effectors of STING pathway inside the bladder cavity. In addition, single IT administration of E7766 to a subcutaneous (SC) tumor in mice bearing dual CT26 tumors in liver and SC lesion cured 90% of animals without recurrence for over 8 months. Those tumor-free animals rejected re-challenge of the same tumor cells in the absence of CD8+ T cells or NK cells, indicating the presence of a highly effective immune memory response following treatment with E7766 independent of either cell population alone. Conclusions: E7766 is a structurally novel and highly potent STING agonist with pan-genotypic activity, demonstrating curative anti-tumoral activity in murine models of BCG-unresponsive NMIBC and of metastatic tumors in deep lesions. Clinical investigation of E7766 is under discussion. Citation Format: Kuan-Chun Huang, Atsushi Endo, Shannon McGrath, Dinesh Chandra, Jiayi Wu, Dae-Shik Kim, Diana Albu, Christy Ingersoll, Karen Tendyke, Kara Loiacono, Thomas Noland, David Verbel, Chi Zhang, Ming-Hong Hao, Mark Matijevic, Vaishali Dixit, Renee R. Hukkanen, Janna Hutz, John Wang, Frank Fang, Xingfeng Bao, Donna Kolber-Simonds, Muzaffar Akram, Nadeem Sarwar. Discovery and characterization of E7766, a novel macrocycle-bridged STING agonist with pan-genotypic and potent antitumor activity through intravesical and intratumoral administration [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3269.

Celecoxib was characterized as a substrate of human cytochrome P450 (CYP) 2D6 in vitro. In recombinant CYP2D6, celecoxib hydroxylation showed atypical substrate inhibition kinetics with apparent Km, Ki, and Vmax of 67.2 μM, 12.6 μM, and 1.33 μM/min, respectively. In human liver microsomes (HLMs), a concentration-dependent inhibition of celecoxib hydroxylation by quinidine was observed after CYP2C9 and CYP3A4 were inhibited. In individual HLMs with variable CYP2D6 activities, a significant correlation was observed between celecoxib hydroxylation and CYP2D6-selective dextromethorphan O-demethylation when CYP2C9 and CYP3A4 activities were suppressed (r = 0.97, P < 0.0001). Molecular modeling showed two predominant docking modes of celecoxib with CYP2D6, resulting in either a substrate or an inhibitor. A second allosteric binding antechamber, which stabilized the inhibition mode, was revealed. Modeling results were consistent with the observed substrate inhibition kinetics. Using HLMs from individual donors, the relative contribution of CYP2D6 to celecoxib metabolism was found to be highly variable and dependent on CYP2C9 genotypes, ranging from no contribution in extensive metabolizers with CYP2C9*1*1 genotype to approximately 30% in slow metabolizers with allelic variants CYP2C9*1*3 and CYP2C9*3*3. These results demonstrate that celecoxib may become a potential victim of CYP2D6-associated drug-drug interactions, particularly in individuals with reduced CYP2C9 activity.

Abstract Chlorotoxin is an established tumor targeting peptide that naturally occurs in scorpion venom. 131I-labelled chlorotoxin assessed in early phase human glioma clinical trials achieved promising results. A peptide drug conjugate (PDC) composed of chlorotoxin peptide linked to a cytotoxic payload (cryptophycin analog) was used as a tool to probe the tumor targeting mechanism of chlorotoxin. The PDC proved efficacious, yet differential sensitivity was observed in multiple human tumor models. Previously described chlorotoxin targets did not align with observed PDC activity; therefore, studies to further elucidate its mechanism were undertaken. PDC treatment of distinct xenograft models led to a wide range of antitumor activity even though similar levels of active metabolite were present in tumor lysates; thus it was hypothesized that a role for chlorotoxin in uptake may be relevant. We identified that the endocytic receptor Neuropilin1 (NRP1) binds to chlorotoxin peptide fragments following proteolytic digestion in vitro. NRP1 binding was selective for peptides with a free C-terminal arginine, while native chlorotoxin which has an amidated C-terminal arginine did not bind. Recovery of chlorotoxin from ex vivo tumor lysate revealed its metabolism to a carboxylated C-terminal arginine version of the peptide, capable of binding to NRP1. These data suggest that chlorotoxin acts as a cryptic peptide incapable of binding to NRP1 systemically and only when metabolized in the tumor microenvironment is NRP1 binding revealed. The expression level of human NRP1 in tumors correlated to the PDC antitumor activity in multiple xenograft models; a wider therapeutic window was observed when NRP1 was highly expressed. Reduction of NRP1 levels in vivo through administration of NRP1 blocking antibodies or by NRP1 knockout in tumor cells blunted PDC antitumor activity while not affecting activity of the cytotoxic payload alone. Reduced PDC antitumor activity correlated with significantly lower levels of active metabolite detected in NRP1-deficient tumors. Together, our findings suggest that chlorotoxin metabolized in the tumor microenvironment binds NRP1 on tumor cells to increase uptake of active metabolite into cells, resulting in enhanced antitumor activity. The identification of NRP1 as an uptake mechanism for chlorotoxin will enable selection of tumors for treatment with chlorotoxin-based therapeutics. Citation Format: Donna Kolber-Simonds, Jiayi Wu, Utpal Majumder, Daniel Custar, Danyang Li, Hong Du, Maarten H. Postema, Thomas Noland, Andrew Hart, George Lai, Sean Eckley, Vaishali Dixit, Karen Tendyke, Kenichi Nomoto, Mary Woodall-Jappe, Sharon McGonigle. Role for neuropilin1 in mode of action of chlorotoxin [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3961.

Long-term coculture models of hepatocytes are promising tools to study drug transport, clearance, and hepatoxicity. In this report we compare the basal expression of drug disposition genes and the inductive response of prototypical inducers (rifampin, phenobarbital, phenytoin) in hepatocyte two-dimensional monocultures and the long-term coculture model (HepatoPac). All the inducers used in the study increased the expression and activity of CYP3A4, CYP2B6 and CYP2C enzymes in the HepatoPac cultures. The coculture model showed a consistent and higher induction of CYP2C enzymes compared with the monocultures. The EC50 of rifampin for CYP3A4 and CYP2C9 was up to 10-fold lower in HepatoPac than the monocultures. The EC50 of rifampin calculated from the clinical drug interaction studies correlated well with the EC50 observed in the HepatoPac cultures. Owing to the long-term stability of the HepatoPac cultures, we were able to directly measure a half-life (t1/2) for both CYP3A4 and CYP2B6 using the depletion kinetics of mRNA and functional activity. The t1/2 for CYP3A4 mRNA was 26 hours and that for the functional protein was 49 hours. The t1/2 of CYP2B6 was 38 hours (mRNA) and 68 hours (activity), which is longer than CYP3A4 and shows the differential turnover of these two proteins. This is the first study to our knowledge to report the turnover rate of CYP2B6 in human hepatocytes. The data presented here demonstrate that the HepatoPac cultures have the potential to be used in long-term culture to mimic complex clinical scenarios.

Benzimidazole 1 is the lead compound resulting from an antibacterial program targeting dual inhibitors of bacterial DNA gyrase and topoisomerase IV. With the goal of improving key drug-like properties, namely, the solubility and the formulability of 1, an effort to identify prodrugs was undertaken. This has led to the discovery of a phosphate ester prodrug 2. This prodrug is rapidly cleaved to the parent drug molecule upon both oral and intravenous administration. The prodrug achieved equivalent exposure of 1 compared to dosing the parent in multiple species. The prodrug 2 has improved aqueous solubility, simplifying both intravenous and oral formulation.